Traffic Route Generation and Adaptation Using Case-Based Reasoning
Andrew J. Whitsitt, Larry E. Travis · Journal of Intelligent Transportation Systems · 1996
Traffic congestion is becoming a severe problem in many large cities, and several Intelligent Transportation Systems (ITS) initiatives have been proposed to address this problem. Providing automated route guidance is an important feature of many of these programs; however, the aspect of providing route guidance based on past experience has received little attention. We propose a system based on case-based reasoning to provide such guidance with minimal recomputation. Case-based reasoning stores instances of previous problem solving and then adapts them to new situations. We describe the architecture and implementation of the system and evaluate the performance of the system on a series of tests. We find that when the system can adapt previously computed routes to solve new problems, solution time is greatly reduced.